Eur J Epidemiol DOI 10.1007/s10654-014-9925-0
PERINATAL EPIDEMIOLOGY
Early pregnancy exposure to antihistamines and risk of congenital heart defects: results of two case–control studies Huberdina P. M. Smedts • Linda de Jonge • Sarah J. G. Bandola • Marlies E. Baardman • Marian K. Bakker • Bruno H. C. Stricker • Re´gine P. M. Steegers-Theunissen
Received: 24 September 2013 / Accepted: 9 June 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract We aimed to study the association between use of antihistamines in early pregnancy and congenital heart defects (CHD) in the offspring. Design: Two case–control studies. Setting: HAVEN study, Erasmus MC, University Medical Centre, Rotterdam, and Eurocat Northern Netherlands (NNL), University Medical Center Groningen, Groningen, the Netherlands. We studied 361 children with CHD and 410 controls without congenital malformations from the HAVEN study and replicated the analyses in 445 children with CHD and 530 controls from the Eurocat NNL registry. Information about antihistamine use in early pregnancy and potential confounders was obtained from questionnaires postpartum. We calculated the association between antihistamines and CHD risk by multivariable logistic regression analysis. Main outcome measures: Odds H. P. M. Smedts S. J. G. Bandola R. P. M. Steegers-Theunissen (&) Department of Obstetrics and Gynaecology, Erasmus MC, University Medical Centre, Dr. Molewaterplein 40, Room Ee 2271a, 3015 GD Rotterdam, The Netherlands e-mail: [email protected] L. de Jonge M. E. Baardman M. K. Bakker Eurocat Registration Northern Netherlands, Department of Genetics, University of Groningen, University Medical Centre Groningen, Groningen, The Netherlands B. H. C. Stricker Department of Epidemiology, Erasmus MC, University Medical Centre, Rotterdam, The Netherlands B. H. C. Stricker Drug Safety Unit, Inspectorate of Healthcare, The Hague, The Netherlands R. P. M. Steegers-Theunissen Department of Clinical Genetics, Erasmus MC, University Medical Centre, Rotterdam, The Netherlands
ratios (OR) with 95 % confidence intervals (CI). In the HAVEN study, 25 of 771 mothers used antihistamines that were associated with an increased CHD risk (OR 3.0, 95 % CI 1.2–7.3), particularly atrioventricular septal defects (AVSD) (OR 5.1, 95 % CI 1.3–20.5) and perimembranous ventricular septal defects (pVSD) (OR 5.1, 95 % CI 1.8–14.4). Mothers with severe nausea who did not use antihistamines had a reduced risk (OR 0.7, 95 % CI 0.5–0.98), whereas nauseous mothers using antihistamines showed an almost fivefold increased risk of pVSD (OR 4.8, 95 % CI 1.1–21.8). The association between antihistamines and AVSD was confirmed in the Eurocat cohort (OR 3.5, 95 % CI 1.4–8.7), but we could not replicate the association with overall CHD risk. We found a positive association between antihistamine use in early pregnancy and CHD risk, particularly AVSD, which seemed to be independent of nausea/vomiting. Keywords Heart anomalies Antihistamine Prevention Teratogen Meclozine Introduction Congenital heart defects (CHD) affect 7.2 per 1,000 live births and occur substantially more often among stillbirths and miscarriages [1, 2]. Due to a high infant morbidity and mortality, CHD also impose a considerable burden of personal suffering and societal costs [3]. Over the past 10 years, epidemiological studies have made major breakthroughs in understanding both the inherited and noninherited causes of CHD [4–6]. However, so far, only around 15 % of CHD can be attributed to a known cause; the majority are thought to result from complex interactions between largely unknown subtle genetic variations and periconceptional exposures [7, 8].
123
H. P. M. Smedts et al.
The thalidomide tragedy of the late 1950s and early 1960s made us aware that medication use in early pregnancy poses serious risks to the fetus [9]. At present, several medicines have been shown to be teratogens for heart development, such as retinoids and anticonvulsants [5]. Since the removal of BendectinÒ (doxylamine/pyridoxine) from the American market in 1983 because of claims of teratogenicity and litigation, pharmacotherapeutic interventions for pregnancy-related nausea/vomiting have been viewed with great suspicion [10]. Nausea/vomiting affect a mean rate of 70 % of pregnant women and the symptoms usually appear during between the 4th and 14th week [11]. Antihistamines for treating pregnancy-related nausea/ vomiting are among the most frequently prescribed medications during pregnancy [12]. In pregnant rats, exposure to antihistamines has been shown to increase the frequency of congenital malformations [13, 14], although evidence from human studies during pregnancy is inconsistent [15–18]. Nonetheless, antihistamines are still popular and widely prescribed during early pregnancy. In Germany, 14 % of pregnant women received anti-emetic prescriptions [19]. CHD is one of the most prevalent congenital malformations developing in the same period that nausea/vomiting presents. We used data from the HAVEN study and from a population selected from the Eurocat Northern Netherlands (NNL) registry to investigate the relationship between maternal use of antihistamines, prescribed for both pregnancy-related nausea/vomiting and other indications, and the risk of specific CHDs.
Methods Study populations HAVEN study The HAVEN study is a population-based, case–control family study to investigate lifestyles, environmental and genetic determinants in the pathogenesis and prevention of CHD [20]. HAVEN is a Dutch acronym for Heart Defects, Vascular status, Genetic factors and Nutrients. 361 eligible cases were identified at the age of around 16 months from the registries of four tertiary referral hospitals in Amsterdam, Leiden and Rotterdam, in the western part of the Netherlands. Two pediatric cardiologists trained at the same hospital diagnosed all the CHD phenotypes after birth by echocardiography and/or cardiac catheterisation and/or surgery. We assume that the different CHD phenotypes may be due to comparable exposures, but that they develop differently depending on genetic background and timeframe of the exposure. The phenotypes we included were: Tetralogy of Fallot (n = 44), complete atrioventricular
123
septal defects (AVSD) (n = 37), perimembranous ventricular septal defect (pVSD) (n = 98), aortic valve stenosis (n = 8), pulmonary valve stenosis (n = 60), coarctation of the aorta (n = 32), transposition of the great vessels (n = 50), hypoplastic left heart syndrome (n = 17), and miscellaneous types (n = 15) consisting of univentricular heart, aorta interruption, mitral valve atresia, tricuspid valve atresia, and corrected transposition of the great arteries. A genetic disorder or syndrome was present in 58/361 of the cases (16.0 %). 410 controls were randomly selected from the registries of the public child healthcare centres covering the source population of the cases. The Dutch healthcare system includes a regular check up of all newborns for health, growth and development by pediatric physicians in such a centre. The controls did not have any major malformations or chromosomal defects according to the medical records obtained at 16 months of age. They were invited to participate at the same age as the cases, at around 16 months. None of the children in the study population had been adopted. Participating families were not related to each other and were able to speak, read and write the Dutch language. We had 74.5 % response from the case families and 61.4 % from the control families. The main reasons not to participate were giving no permission to take a blood sample from their child for research purposes and the expected time effort. The study design has been described in more detail previously [20]. The Central Committee of Research in Humans and the Medical Ethical Committees of the participating hospitals approved the protocol and written consent from the parents was obtained prior to their participation. For the analyses, we included 361 cases with CHD and 410 controls between October 2003 and February 2007. Mothers with diabetes (n = 6), hyperhomocysteinaemia (n = 1), epilepsy (n = 6), hypertension (n = 6) and rheumatoid arthritis (n = 2) were excluded as these conditions could confound the associations we were investigating. We chose to study the mother–child pairs at 16 months after birth; this time of investigation was standardised and minimises recall bias and differential effects between the recall by case and control mothers regarding their periconceptional medication use, nausea/vomiting, and lifestyle behaviours. In addition, the study moment at more than 1 year after birth was chosen because most CHD are detected within the first year of life, thereby minimizing misclassification of control children. The mothers completed a questionnaire on demographics, illnesses, lifestyle behaviors and medication use at the study moment and during their periconceptional period, defined as 3 months prior to conception and up to 10 weeks after conception. The questionnaires were filled in at home and checked for
Results of two case–control studies
completeness and consistency by the researcher during the participant’s visit to our clinic. The history of prescribed and over the counter antihistamine use during the first 10 weeks of pregnancy was obtained from the questionnaire and coded according to the internationally accepted Anatomical Therapeutic Chemical (ATC) classification (controlled by the World Health Organisation Collaborating Centre for Drug Statistics Methodology) [21]. Medicines were coded based on their pharmacological and chemical properties. Anti-emetics were meclozine, meclozine/pyridoxine combination and cyclizine (piperazines), other antihistamines were cetirizine and levoceterizine (piperazines), amino-alkyl ethers and phenothiazines. We defined a child as having been exposed to the medicine during the first 10 weeks of pregnancy if the mother had taken the medicine during this 10-week period. In addition, the history of pregnancy-related nausea/vomiting was evaluated. Information on nausea/vomiting in the first 10 weeks of pregnancy was categorised as no nausea or vomiting, mild nausea with or without incidental vomiting, and severe nausea/vomiting. The last category was characterised by daily nausea with vomiting or with a serious influence on dietary intake. Mothers who reported any use of alcohol or cigarettes were considered as alcohol users and smokers. Periconceptional vitamin use was defined as daily intake during the whole periconceptional period, defined as starting 3 months prior to conception and up to 10 weeks after conception. Educational level was classified according to the definitions of Statistics Netherlands into low (primary/lower, vocational/intermediate, secondary), intermediate (intermediate, vocational/higher, secondary) and high education (higher vocational/university) [22]. Ethnicity was categorised as Western (both parents were born in the Netherlands, or in a European country or non-Western (one of the parents was of non-European origin) [22]. Standardised maternal measurements of weight (weighing scale, SECA, Germany, accurate to 0.5 kg) and height (anthropometric rod, SECA, Germany, accurate to 0.1 cm) were also performed during the clinic visit at 16 months after delivery. Body mass index (BMI) after birth was defined as weight in kilograms divided by the height in meters squared. Eurocat Northern Nederlands (NNL) Eurocat NNL is a population-based birth defect registry in the northern part of the Netherlands (the three Northern provinces Friesland, Groningen and Drenthe); it covers approximately 10 % of all Dutch births. A child can be registered with a defect up to the age of 16, there is no lower age limit. All types of births are included in the registry: live births, stillbirths, spontaneous abortions and terminations.
Parental informed consent is required to register a child and parents are asked to fill in a questionnaire on sociodemographic characteristics, prenatal screening methods and diagnostic tests, and prenatal exposure to possible risk factors (chemicals, drugs, etc.). With consent of the mother information on prescribed medications from the period 3 months before conception up to delivery are obtained from her pharmacy. The actual use of the prescribed medication and use of over the counter medication is later verified in a telephone interview. Birth defects are coded according to the ICD coding system. For births up to 2001, ICD-9 is used, while for births from 2002 onwards, ICD-10 is used [21]. Medication taken by the mother is coded according to the ATC classification system. From the Eurocat NNL database, 445 cases and 530 controls were selected. All the children were born in a 12-year period between 1st January 1997 and 31st December 2008. To comply with the HAVEN study, only live births were included. Of the 445 cases, 58 were diagnosed with Tetralogy of Fallot, 58 with complete AVSD, 121 pVSD, 14 with aortic valve stenosis, 39 with pulmonary valve stenosis, 51 with coarctation of the aorta, 66 with transposition of the great vessels and 38 with hypoplastic left heart syndrome. In total, 74/445 of the Eurocat cases (16.6 %) also had a genetic disorder. Since Eurocat NNL only includes children and fetuses with a birth defect, children with isolated hip dysplasia and/or luxation were selected as controls. The use of antihistamines has not been associated with these disorders. We excluded mothers with diabetes, hyperhomocysteinaemia, epilepsy, hypertension and rheumatoid arthritis because these conditions could confound the associations we were investigating. Anti-emetics were meclozine, meclozine/pyridoxine combination and cyclizine (piperazines), other antihistamines were cetirizine and levoceterizine (piperazines), amino-alkyl ethers and phenothiazines. Exposure to an antihistamine was defined as any antihistamine use during the first 14 weeks of pregnancy, calculated from the first day of the last menstrual period. No information was available on nausea/vomiting. Mothers who reported any periconceptional use of alcohol and/or cigarettes were considered as alcohol users and/or smokers. Periconceptional use of folic acid or multivitamins containing folic acid was defined as any use of these during the 3 months before conception up to the end of the first trimester. To record educational level, we used the definitions set by Statistics Netherlands. Ethnicity was categorised as ‘Western’ if the mother was born in Europe, North America, Australia, New Zealand or Indonesia, and ‘non-Western’ if the mother was born in any another country. Maternal weight and height
123
H. P. M. Smedts et al.
were based on the situation before pregnancy and were self-reported. Statistical analyses HAVEN study In the HAVEN analyses, normality of continuous variables was tested by the one-sample Kolmogorov–Smirnov Test. BMI showed a positively skewed distribution even after logarithmic transformation. Continuous variables of the maternal characteristics were therefore presented as medians with interquartile ranges, and differences between cases and controls were tested by the Mann–Whitney U test. Categorical variables were tested between cases and controls by the Chi squared test. We used a forward, stepwise, multivariable regression model to study associations between use of antihistamines during early pregnancy, stratified for anti-emetics and other antihistamines, and the risk of both overall CHD and the separate CHD phenotypes. The risk estimates were adjusted for the variables that were either significantly different between cases and controls or that had a P \ 0.1 in the logistic model (maternal age, educational level, ethnicity, parity, sex of the child and any medication except antihistamines). Crude and adjusted odds ratios (ORs) and 95 % confidence intervals (CIs) were calculated for use of antihistamines during early pregnancy if at least two cases had been exposed. In the HAVEN data, we then performed a combined analysis of anti-emetic use in early pregnancy and nausea/vomiting, with no antihistamine use and no nausea/ no vomiting as a reference, and separately presented the results for complete AVSD and pVSD defects, as these were found to be associated with antihistamine use. Trends across severity of nausea/vomiting towards CHD risk were evaluated by the linear-by-linear association test. Statistical analyses were performed with SPSS for Windows software (version 20.0; SPSS Inc, Chicago, IL, USA). Eurocat NNL The normality of continuous variables, maternal age and BMI was tested by the one-sample Kolmogorov–Smirnov Test. Maternal age was normally distributed for the Eurocat population. These characteristics were therefore presented as mean and standard deviation, and the differences between cases and controls were tested by the independent t test. Since BMI was not normally distributed, these characteristics were presented as medians with interquartile ranges and the differences between cases and controls were tested by the Mann–Whitney U test. Categorical variables were tested between cases and controls by the Chi squared
123
test. We used a logistic regression model to study associations between early pregnancy use of antihistamine, stratified for anti-emetics and other antihistamines, and the risk of both overall CHD and the separate CHD phenotypes. The risk estimates were adjusted for the same potential confounders or effect modifiers as in the HAVEN study. Crude and adjusted ORs and 95 % CIs were calculated if at least two cases had been exposed.
Results Socio-demographic and lifestyle characteristics of case mothers of a child with CHD and controls are presented in Table 1. HAVEN study case mothers showed a slightly higher age than controls, but ethnicity, educational level, BMI, parity, child gender and periconceptional exposures were not different except for a higher medication use in case mothers (27 %) than in healthy controls (19 %). For the Eurocat population case mothers were also slightly older than control mothers. The large majority of cases and controls were of Western origin, however the cases were slightly more often of non-Western origin (3 % [15/445] vs. 1 % [5/530]). Furthermore, we found a statistically significant difference in educational level and parity between cases and controls. There were more girls among the controls, since hip dysplasia is more common among girls. In the HAVEN study, antihistamine use was reported in 3.2 % of case and control mothers, of which 5.0 % in case mothers (18 of 361) and 1.7 % in control mothers (7 of 410) (Table 2). The use of antihistamines was associated with a threefold increased risk of CHD (crude OR 3.0, 95 % CI 1.2–7.3). Because of the small number of exposed cases, adjustment for potential confounders might not be useful. However, for completeness we added the adjusted ORs to Table 2. After stratification for anti-emetics, i.e., meclozine and the meclozine/pyridoxine combination as well as other antihistamines, the risk estimates were comparable although significance was lost due to small numbers. Stratification per CHD phenotype in the HAVEN study revealed that antihistamine use, particularly antiemetic medication, was associated with the occurrence of complete AVSD (crude OR 5.8, 95 % CI 1.03–32.8) and pVSD (crude OR 5.5, 95 % CI 1.4–20.7). Furthermore, other antihistamines were associated with an increased risk of Tetralogy of Fallot (crude OR 6.5, 95 % CI 1.05–39.8). Risk estimates for coarctation of the aorta, aortic valve stenosis, pulmonary valve stenosis, hypoplastic left heart syndrome, and the miscellaneous subgroup were not calculated as fewer than two cases in each category were exposed.
Results of two case–control studies Table 1 Characteristics of mothers of a child with congenital heart defects and of controls
HAVEN
Eurocat
CHD (n = 361)
Controls (n = 410)
CHD (n = 445)
Controls (n = 530)
32.9 (29.8–36.2)*
32.7 (28.8–35.1)
31.4 (4,543)^
30.0 (4,032)
24.4 (22.0–28.0)
24.4 (22.1–28.0)
23.5 (21.4–26.6)
23.5 (21.4–25.9)
300 (83)
318 (78)
422 (97)#
518 (99)
61 (17)
92 (22)
15 (3)
5 (1)
Low Intermediate
90 (25) 161 (45)
96 (24) 199 (48)
91 (21)* 211 (49)
75 (16) 280 (59)
High
Maternal age (years) 2
BMI (kg/m ) Ethnicity Western Non-Western Educational level HAVEN controls are healthy children without congenital malformation. Eurocat controls are children with an isolated hip disorder (dysplasia or dislocation). Data are median (interquartile range), mean (standard deviation) or n (%), compared by Mann–Whitney U Test or Chi squared test
110 (30)
115 (28)
126 (29)
124 (26)
Parity (first child)
152 (42)
200 (49)
147 (34)^
249 (48)
Sex of the child (male)
205 (57)
229 (56)
256 (58)^
78 (15)
Periconceptional use of 138 (38)
132 (32)
88 (21)
86 (17)
Cigarettes, yes
63 (18)
89 (22)
114 (26)
108 (21)
CHD congenital heart defect
B-vitamins, yes
182 (50)
206 (50)
300 (67)
367 (69)
*
Any medication, yes
98 (27)#
78 (19)
191 (43)
203 (38)
Antihistamines, yes
18 (5)#
7 (2)
22 (5)
20 (4)
P value B 0.05, # P value B 0.01, ^ P value B 0.001
Alcohol yes
In the Eurocat population, antihistamine use was reported in 4.3 % of case and control mothers, of which 4.9 % in case mothers (22/445) and 3.8 % in control mothers (20/530). The positive association between periconception use of antihistamine medication and complete AVSD was confirmed (OR 3.5, 95 % CI 1.4–8.7) in the Eurocat population. However, the association between periconception use of antihistamine medication and the total group of CHDs, pVSD and Tetralogy of Fallot was not confirmed: respectively we found OR 1.3, 95 % CI 0.7–2.5; OR 0.9, 95 % CI 0.3–2.6 and OR 0.9, 95 % CI 0.2–4.1. In the HAVEN study, in a subgroup analysis after excluding cases with a known genetic factor, the observed associations remained between antihistamine use and overall CHD risk (OR 3.2, 95 % CI 1.3–8.0), AVSD (OR 11.5, 95 % CI 1.2–111) and pVSD (OR 6.0, 95 % CI 2.0–17.7). In the Eurocat population, the association between antihistamines other than anti-emetics and AVSD also remained after excluding cases with a known genetic factor (OR 15.9, 95 % CI 1.4–184), but the association between overall antihistamine use and AVSD (OR 3.0, 95 % CI 0.7–14.0) attenuated to non-significant. In the HAVEN study, 52 % of case mothers and 59 % of control mothers reported nausea/vomiting. We demonstrated a significant trend towards a reduced overall CHD risk by increasing severity of nausea/vomiting, P = 0.020. Severe nausea and vomiting without the use of anti-emetic medication seemed to be associated with a reduced risk of CHD (OR 0.7, 95 % CI 0.5–0.98) (Table 3). However,
mothers with severe nausea/vomiting who used anti-emetic medication tended to have an increased risk of overall CHD offspring, albeit non-significant. Remarkably, the reduced risk in the group of mothers with severe nausea/ vomiting changed into an almost fivefold increased risk for particularly pVSD (OR 4.8, 95 % CI 1.1–21.8). These analyses could not be replicated as this information was not available for the Eurocat population.
Discussion Main findings The HAVEN study showed that antihistamine use was associated with a threefold increased overall CHD risk and a fivefold increased risk of pVSD and AVSD. In addition, a positive association was found with Tetralogy of Fallot for antihistamines other than anti-emetics. These associations were independent of the mother’s nausea/vomiting. In the Eurocat population we confirmed the association between the mother’s antihistamine use and the increased risk of having a child with AVSD, but we could not confirm the other associations. Strengths and limitations One of the strong points of our study is that we replicated the association with AVSD in two large, independent
123
123 3/7 8/7
Tetralogy of Fallot, n = 44 Transposition of the great vessels, n = 50
Complete AVSD, n = 37
Perimembranous VSD, n = 98
3/20 7/20 4/20
Tetralogy of Fallot, n = 58
Transposition of the great vessels, n = 66
Complete AVSD, n = 58
Perimembranous VSD, n = 121
3.0 (1.2–7.3)
0.9 (0.3–2.6)
3.5 (1.4–8.7)
1.2 (0.4–4.2)
0.9 (0.2–4.1)
1.3 (0.7–2.5)
5.1 (1.8–14.4)
5.1 (1.3–20.5)
2.7 (0.6–13.6) 2.4 (0.5–11.9)
2.6 (1.02–6.7)
0.7 (0.2–2.3)
3.6 (1.3–9.9)
0.9 (0.2–3.9)
0.4 (0.1–2.8)
1.2 (0.6–2.5)
5.2 (1.6 –16.7)
9.0 (1.6–50.8)
3.4 (0.6–19.9) 2.4 (0.4–14.2)
10/4
4/18
5/18
3/18
2/18
20/18
5/4
2/4
0/4 1/4
2.9 (0.9–9.3)
1.0 (0.3–2.9)
2.8 (1.0–7.8)
1.3 (0.4–4.7)
1.0 (0.2–4.5)
1.3 (0.7–2.6)
5.5 (1.4–20.7)
5.8 (1.03–32.8)
NE NE
0.8 (0.2–2.8)
2.8 (0.9–9.0)
1.3 (0.3–5.4)
0.5 (0.1–3.8)
1.2 (0.5–2.6)
4.5 (1.1–18.8)
6.8 (0.9–50.7)
NE NE
2.2 (0.6–7.2)
OR (95 % CI)*
0/2
2/2
0/2
0/2
2/2
3/3
1/3
2/3 1/3
8/3
Cases/ controls
NE
10,0 (1.4–72.4)
NE
NE
1.2 (0.2–8.6)
4.3 (0.9–21.6)
NE
6.5 (1.05–39.8) NE
3.1 (0.8–11.7)
OR (95 % CI)
Other antihistamines
NE
9.4 (1.0–85.2)
NE
NE
1.2 (0.1–10.1)
4.3 (0.8–23.7)
NE
8.5 (1.2–63.2) NE
2.8 (0.7–11.2)
OR (95 % CI)*
* Odds ratios adjusted to maternal age, and educational level, ethnicity, parity, sex of the child and periconceptional use of any medication except antihistamines. Anti-emetics are meclozine, meclozine/pyridoxine combination and cyclizine (piperazines). Other antihistamines are cetirizine and levoceterizine (piperazines), amino-alkyl ethers and phenothiazine
AVSD atrioventricular septal defect, VSD ventricular septal defect, CHD congenital heart defect, OR odds ratio, CI confidence interval, NE not estimated from logistic regression model
22/20 2/20
Total CHD, n = 445
Eurocat database, n = 445/530
18/7 2/7 2/7
Total CHD, n = 361
HAVEN database, n = 361/410
OR (95 % CI)
Cases/ controls
OR (95 % CI)*
Cases/ controls
OR (95 % CI)
Anti-emetics
Total antihistamines
Table 2 Early pregnancy exposure to antihistamines, stratified for anti-emetics and other antihistamines, and the risk of congenital heart defects
H. P. M. Smedts et al.
AVSD Atrioventricular septal defect, VSD ventricular septal defect, CHD congenital heart defect, OR odds ratio, CI confidence interval, NE not estimated from logistic regression model
* Odds ratios adjusted to maternal age, and educational level, ethnicity, parity, sex of the child, periconceptional use of any medication except antihistamines. The reference group comprised all mothers not exposed to antihistamine medication and who had no complaints of nausea and vomiting. We excluded 8 cases and 3 controls using antihistamines other than anti-emetics for this analysis
NE
4.8 (1.1–21.6) 5.4 (1.4–21.0) 5/4 6.8 (0.9–53.2) 5.2 (0.9–30.1) 2/4 1.8 (0.5–6.1) 10/4 ?
Severe
2.5 (0.8–8.2)
NE NE
NE 0/0
0/0 NE
NE NE
NE 0/0
0/0 NE
NE
0/0
NE
0/0
Mild ?
?
No
NE
1.0 (0.5–1.7)
1.0 (0.6–1.8) 1.0 (0.6–1.8)
0.8 (0.5–1.5) 23/117
28/118 1.0 (0.4–2.3)
0.9 (0.3–2.2) 0.7 (0.3–1.7)
0.9 (0.4–2.0) 10/118
8/117 0.7(0.5–1.01)
0.8(0.6–1.1) 0.8 (0.6–1.2)
Severe
0.7 (0.5–0.98)
96/118
80/117
Mild
-
Reference 39/168 Reference Reference Reference 167/168
(n = 36/407)
OR (95 % CI)* OR (95 % CI)
16/168
Reference
OR (95 % CI) Cases/controls
Cases/ controls (n = 353/ 407)
No -
-
Reference
OR (95 % CI)* OR (95 % CI) Cases/ controls (n = 95/407) OR (95 % CI)*
Perimembranous VSD Complete AVSD Total CHD Nausea/vomiting Anti-emetic use
Table 3 Maternal use of anti-emetics in combination with nausea and vomiting during early pregnancy in association with risk of congenital heart defect, HAVEN study results
Results of two case–control studies
populations with good information about medication use during pregnancy. In the HAVEN study, results from a standardised questionnaire were verified in a personal interview, while in the Eurocat registry, information on medication use is based on pharmacy records and verified in telephone interviews. Another strength of our study is the accuracy of the diagnoses of the specific CHD phenotypes: for the HAVEN study, two pediatric cardiologists diagnosed all the CHD cases, while for Eurocat, a medical doctor and clinical geneticist specializing in heart anomalies classified the cases and controls. One weakness of our study is that there could have been selection bias since only live births were included. CHD can be part of a chromosomal or genetic disorder and such pregnancies are terminated relatively more often or result in early fetal loss. Another type of bias that always has to be considered in case–control studies is recall bias [23]. In the HAVEN study we therefore standardised the data collection shortly after pregnancy. In Eurocat, the pharmacy data was for prescriptions dispensed shortly before and during pregnancy, and their use was verified with the mother. Use of anti-emetics was 5.0 and 5.4 % in the HAVEN and Eurocat populations, respectively, which agrees with the 5.8 % reported in a study on drug prescription patterns in the Netherlands [12]. Frequencies of use by controls were, however, lower at 1.7 and 3.9 %. No other medicines or maternal illnesses could explain the difference in overall medication use between cases and controls. If, in the control group, under-reporting is an issue, differential recall bias cannot be excluded in the HAVEN study. However, if there was selective recall bias, we would have expected mothers of a CHD child to recall more nausea/vomiting for which they used medication than the controls. The mothers were not aware of our specific questions on associations between medication use and CHD. In the HAVEN study, a fixed 2 year time interval was chosen between first trimester of pregnancy and the interview, which was similar between cases and controls. The distance between the region of Eurocat NNL and the Western part of the Netherlands where the HAVEN study cases were recruited is significant. Therefore, there is only a very small risk of overlap in patients between the two databases. Finally, we are aware that our observed associations are based on a small number of antihistamine exposures, leading to imprecise risk estimates with large confidence intervals. Interpretation Our findings are in line with the results reported by Queißer-Luft et al. [15] who showed that early pregnancy use of anti-allergics, mainly antihistamines, was associated with a seven- to nine-fold increased risk of CHD and
123
H. P. M. Smedts et al.
musculoskeletal anomalies. Our results might underestimate the true risk if antihistamines are also associated with cases of isolated hip dysplasia, which were used as controls in the Eurocat analyses. Recent data from the National Birth Defect Prevention Study also revealed positive associations between doxylamine, which is the major compound of BendectinÒ, and spina bifida, cleft lip, left ventricular outflow tract obstruction defects, and hypoplastic left heart syndrome [24]. Associations were also demonstrated between meclozine use and orofacial clefts, and between antihistamine use and pVSDs. In a prospective cohort study among 16,536 women exposed to meclozine and 540,660 unexposed women, there was no difference in the occurrence of congenital malformations. Although this was a large cohort, the data may be distorted by confounding by indication, since their results were not adjusted for nausea/vomiting [16]. In a second prospective cohort study, investigating frequencies of congenital malformations in 196 women with first trimester exposure to cetirizine and 1,686 controls, no increased risk was found for any congenital malformation, but the study might have been underpowered and medication among control mothers was not specified [17]. Furthermore, the number of exposed cases is too limited to detect only a modest teratogenic effect of antihistamine use. A meta-analysis revealed that antihistamine use in early pregnancy protected against all types of major congenital malformations [18]. However, by pooling all types of malformations, a differential effect for a specific birth defect would be considerably diluted. Moreover, they did not investigate confounding by indication. A meta-analysis of 16 cohort and 11 case–control studies on the association with BendectinÒ, which is pharmacologically closely related to meclozine, showed no difference in risk [25]. It is conceivable that the presence of nausea/vomiting indicates the adaptation of the maternal endocrine and metabolic system to pregnancy, and as such may protect the embryo from harmful or teratogenic exposures [26]. However, severe nausea/vomiting has been associated with orofacial clefting, renal dysgenesia and urinary tract defects [27].The absence of nausea/vomiting in pregnant women may also reflect a reduced production of placental human chorionic gonadotrophin and thyroxin, resulting in impaired placental growth and subsequently slower fetal growth and development [28]. Our results suggest that antihistamine medication might be a cardiac teratogen. An alternative hypothesis is that it is not the medication, but rather the resulting reduction in nausea/vomiting that increases CHD risk. A reduction in nausea/vomiting during early pregnancy might result in increased fetal exposure to harmful agents or unknown teratogens that stimulate the mother’s vomiting.
123
The underlying mechanisms for a potential teratogenic effect from antihistamines are unknown and we can therefore only speculate. In adults, blockage of H1-receptors, expressed in several tissues such as heart, placenta, and endothelium, can lead to reflex tachycardia, ventricular arrhythmias and hemodynamic changes in the heart [29]. H1-receptors play an important role in embryonic development [30]. If we hypothesise that the H1-receptors are also expressed in the fetal heart, its inhibition might lead to hemodynamic changes and subsequent CHD. AVSD is strongly correlated with Down syndrome [31]. In the HAVEN study, in total 21 of 33 cases with AVSD, and 2 of 3 exposed AVSD cases, also had Down syndrome. Down syndrome was not correlated with antihistamine use nor with complaints of nausea in the first trimester. In the Eurocat population, 35 of 58 cases with AVSD, and 4 of 7 exposed AVSD cases, had Down syndrome. We did not exclude cases with a genetic abnormality from our analyses, as this is clearly not the only factor determining risk for CHD. In this group, environmental exposures are also important modifiers. Furthermore, in our study population, there may be more cases with an underlying genetic factor, but they cannot be excluded because most of the genetic causes of CHD are not yet known. However, excluding cases with a known genetic factor did not substantially change the results.
Conclusion We found a positive association between maternal antihistamine use and risk of CHD, in particular pVSD and AVSD, in the HAVEN study. It appears that it is not the pregnancy-related nausea/vomiting, but rather the exposure to antihistamines that seems to increase the risk of CHD. The Eurocat data confirmed the association between antihistamine use and AVSD, but not with overall CHD risk; nor could we investigate the association with nausea/ vomiting. Our findings therefore warrant further investigation, which should take into account the need for a high number of exposed cases with a specific CHD phenotype. Acknowledgments We thank Professor W.A. Helbing, MD, PhD, Professor J. Ottenkamp, MD, PhD, and F.M.H. Siebel, MD, for their assistance in recruiting the case and control families. We are also grateful to Dr. M. Wildhagen, for data management and to Dr. A.C. Verkleij-Hagoort, Dr. L.M.J.W. van Driel and S. Borst for help in the data collection. We thank JL Senior for editing the final manuscript. The HAVEN study was financially supported by the Bo Hjelt Foundation (Grant 2005) and the Netherlands Heart Foundation (Grant 2002.B027). Eurocat NNL is funded by the Dutch Ministry of Health, Welfare and Sports. Conflict of interest
There are no conflicts of interest to declare.
Results of two case–control studies Ethical Standard The study protocol was approved by the Central Committee of Research in Humans in The Hague, the Netherlands, and by the local Medical Ethics Committees of the four participating hospitals (CCMO07.1052/MA/P03.0200; MEC212.508/2002/91).
References 1. Hoffman JI, Kaplan S. The incidence of congenital heart disease. J Am Coll Cardiol. 2002;39:1890–900. 2. Dolk H, Loane M, Garne E. Congenital heart defects in Europe: prevalence and perinatal mortality, 2000–2005. Circulation. 2011;123:841–9. 3. Yoon PW, Olney RS, Khoury MJ, Sappenfield WM, Chavez GF, Taylor D. Contribution of birth defects and genetic diseases to pediatric hospitalizations. A population-based study. Arch Pediatr Adolesc Med. 1997;151:1096–103. 4. Pierpont ME, Basson CT, Benson DW Jr, Gelb BD, Giglia TM, Goldmuntz E, et al. Genetic basis for congenital heart defects: current knowledge: a scientific statement from the American Heart Association Congenital Cardiac Defects Committee, Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics. Circulation. 2007;115:3015– 38. 5. Jenkins KJ, Correa A, Feinstein JA, Botto L, Britt AE, Daniels SR, et al. Non-inherited risk factors and congenital cardiovascular defects: current knowledge: a scientific statement from the American Heart Association Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics. Circulation. 2007;115:2995–3014. 6. Rosenquist TH. Folate, homocysteine and the cardiac neural crest. Dev Dyn. 2013;242:201–18. 7. Botto LD, Correa A. Decreasing the burden of congenital heart anomalies: an epidemiologic evaluation of risk factors and survival. Progr Paediatr Cardiol. 2003;18:111–21. 8. Steegers-Theunissen RP, Steegers EA. Nutrient-gene interactions in early pregnancy: a vascular hypothesis. Eur J Obstet Gynecol Reprod Biol. 2003;106:115–7. 9. Dally A. Thalidomide: was the tragedy preventable? Lancet. 1998;351:1197–9. 10. Leeder JS, Spielberg SP, MacLeod SM. Bendectin: the wrong way to regulate drug availability. Can Med Assoc J. 1983;129: 1085–7. 11. Einarson TR, Piwko C, Koren G. Quantifying the global rates of nausea and vomiting of pregnancy: a meta analysis. J Popul Ther Clin Pharmacol. 2013;20:e171–83. 12. Bakker MK, Jentink J, Vroom F, Van Den Berg PB, De Walle HE, De Jong-Van Den Berg LT. Drug prescription patterns before, during and after pregnancy for chronic, occasional and pregnancy-related drugs in the Netherlands. BJOG. 2006;113: 559–68. 13. Narrod SA, Wilk AL, King CT. Metabolism of meclizine in the rat. J Pharmacol Exp Ther. 1965;147:380–4. 14. King CT, Weaver SA, Narrod SA. Antihistamines and teratogenicity in the rat. J Pharmacol Exp Ther. 1965;147:391–8.
15. Queißer-Luft A, Eggers I, Stolz G, Kieninger-Baum D, Schlaefer K. Serial examination of 20,248 newborn fetuses and infants: correlations between drug exposure and major malformations. Am J Med Genet. 1996;63:268–76. 16. Kallen B, Mottet I. Delivery outcome after the use of meclozine in early pregnancy. Eur J Epidemiol. 2003;18:665–9. 17. Weber-Schoendorfer C, Schaefer C. The safety of cetirizine during pregnancy. A prospective observational cohort study. Reprod Toxicol. 2008;26:19–23. 18. Seto A, Einarson T, Koren G. Pregnancy outcome following first trimester exposure to antihistamines: meta-analysis. Am J Perinatol. 1997;14:119–24. 19. Markl GE, Strunz-Lehner C, Egen-Lappe V, Hasford J. Prescribing patterns of anti-emetic drugs during.pregnancy in Germany. Arch Gynecol Obstet. 2007;275:461–7. 20. Verkleij-Hagoort AC, Verlinde M, Ursem NT, Lindemans J, Helbing WA, Ottenkamp J, et al. Maternal hyperhomocysteinaemia is a risk factor for congenital heart disease. BJOG. 2006; 113:1412–8. 21. WHO Collaborating Centre for Drug Statistics Methodology. http://www.whocc.no/atcddd/. Accessed Sept 24 2008. http:// www.icd9data.com/ and http://www.icd10data.com/ICD10CM/ Codes. 22. Statistics Netherlands. Classification of educational level. Voorburg/Heerlen, the Netherlands, 2006. http://www.cbs.nl/menu/ methoden/classificaties/overzicht/soi/default.htm. Accessed Nov 27, 2007. 23. Cogswell ME, Bitsko RH, Anderka M, Caton AR, Feldkamp ML, Hockett Sherlock SM, et al. Control selection and participation in an ongoing, population-based, case-control study of birth defects: the National Birth Defects Prevention Study. Am J Epidemiol. 2009;170:975–85. 24. Gilboa SM, Strickland MJ, Olshan AF, Werler MM, Correa A. The National Birth Defects Prevention Study. Use of antihistamine medications and isolated major malformations. Birth Defect Res Part A. 2009;85:137–50. 25. McKeigue PM, Lamm SH, Linn S, Kutcher JS. Bendectin and birth defects: I. A meta-analysis of the epidemiologic studies. Teratology. 1994;50:27–37. 26. Flaxman SM, Sherman PW. Morning sickness: a mechanism for protecting mother and embryo. Q Rev Biol. 2000;75:113–48. 27. Czeizel AE, Puho´ E, Acs N, Ba´nhidy F. Inverse association between severe nausea and vomiting in pregnancy and some congenital abnormalities. Am J Med Genet A. 2006;140:453–62. 28. Huxley RR. Nausea and vomiting in early pregnancy: its role in placental development. Obstet Gynecol. 2000;95:779–82. 29. Simons FE. Advances in H1-antihistamines. N Engl J Med. 2004;351:2203–17. 30. Matsuyama K, Ichikawa T, Nitta Y, Ikoma Y, Ishimura K, Horio S, et al. Localized expression of histamine H1 receptors in syncytiotrophoblast cells of human placenta. J Pharmacol Sci. 2006; 102:331–7. 31. Marino B, Vairo U, Corno A, Nava S, Guccione P, Calabro R, et al. Atrioventricular canal in Down syndrome. Prevalence of associated cardiac malformations compared with patients without Down syndrome. Am J Dis Child. 1990;144:1120–2.
123
PERINATAL EPIDEMIOLOGY
Early pregnancy exposure to antihistamines and risk of congenital heart defects: results of two case–control studies Huberdina P. M. Smedts • Linda de Jonge • Sarah J. G. Bandola • Marlies E. Baardman • Marian K. Bakker • Bruno H. C. Stricker • Re´gine P. M. Steegers-Theunissen
Received: 24 September 2013 / Accepted: 9 June 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract We aimed to study the association between use of antihistamines in early pregnancy and congenital heart defects (CHD) in the offspring. Design: Two case–control studies. Setting: HAVEN study, Erasmus MC, University Medical Centre, Rotterdam, and Eurocat Northern Netherlands (NNL), University Medical Center Groningen, Groningen, the Netherlands. We studied 361 children with CHD and 410 controls without congenital malformations from the HAVEN study and replicated the analyses in 445 children with CHD and 530 controls from the Eurocat NNL registry. Information about antihistamine use in early pregnancy and potential confounders was obtained from questionnaires postpartum. We calculated the association between antihistamines and CHD risk by multivariable logistic regression analysis. Main outcome measures: Odds H. P. M. Smedts S. J. G. Bandola R. P. M. Steegers-Theunissen (&) Department of Obstetrics and Gynaecology, Erasmus MC, University Medical Centre, Dr. Molewaterplein 40, Room Ee 2271a, 3015 GD Rotterdam, The Netherlands e-mail: [email protected] L. de Jonge M. E. Baardman M. K. Bakker Eurocat Registration Northern Netherlands, Department of Genetics, University of Groningen, University Medical Centre Groningen, Groningen, The Netherlands B. H. C. Stricker Department of Epidemiology, Erasmus MC, University Medical Centre, Rotterdam, The Netherlands B. H. C. Stricker Drug Safety Unit, Inspectorate of Healthcare, The Hague, The Netherlands R. P. M. Steegers-Theunissen Department of Clinical Genetics, Erasmus MC, University Medical Centre, Rotterdam, The Netherlands
ratios (OR) with 95 % confidence intervals (CI). In the HAVEN study, 25 of 771 mothers used antihistamines that were associated with an increased CHD risk (OR 3.0, 95 % CI 1.2–7.3), particularly atrioventricular septal defects (AVSD) (OR 5.1, 95 % CI 1.3–20.5) and perimembranous ventricular septal defects (pVSD) (OR 5.1, 95 % CI 1.8–14.4). Mothers with severe nausea who did not use antihistamines had a reduced risk (OR 0.7, 95 % CI 0.5–0.98), whereas nauseous mothers using antihistamines showed an almost fivefold increased risk of pVSD (OR 4.8, 95 % CI 1.1–21.8). The association between antihistamines and AVSD was confirmed in the Eurocat cohort (OR 3.5, 95 % CI 1.4–8.7), but we could not replicate the association with overall CHD risk. We found a positive association between antihistamine use in early pregnancy and CHD risk, particularly AVSD, which seemed to be independent of nausea/vomiting. Keywords Heart anomalies Antihistamine Prevention Teratogen Meclozine Introduction Congenital heart defects (CHD) affect 7.2 per 1,000 live births and occur substantially more often among stillbirths and miscarriages [1, 2]. Due to a high infant morbidity and mortality, CHD also impose a considerable burden of personal suffering and societal costs [3]. Over the past 10 years, epidemiological studies have made major breakthroughs in understanding both the inherited and noninherited causes of CHD [4–6]. However, so far, only around 15 % of CHD can be attributed to a known cause; the majority are thought to result from complex interactions between largely unknown subtle genetic variations and periconceptional exposures [7, 8].
123
H. P. M. Smedts et al.
The thalidomide tragedy of the late 1950s and early 1960s made us aware that medication use in early pregnancy poses serious risks to the fetus [9]. At present, several medicines have been shown to be teratogens for heart development, such as retinoids and anticonvulsants [5]. Since the removal of BendectinÒ (doxylamine/pyridoxine) from the American market in 1983 because of claims of teratogenicity and litigation, pharmacotherapeutic interventions for pregnancy-related nausea/vomiting have been viewed with great suspicion [10]. Nausea/vomiting affect a mean rate of 70 % of pregnant women and the symptoms usually appear during between the 4th and 14th week [11]. Antihistamines for treating pregnancy-related nausea/ vomiting are among the most frequently prescribed medications during pregnancy [12]. In pregnant rats, exposure to antihistamines has been shown to increase the frequency of congenital malformations [13, 14], although evidence from human studies during pregnancy is inconsistent [15–18]. Nonetheless, antihistamines are still popular and widely prescribed during early pregnancy. In Germany, 14 % of pregnant women received anti-emetic prescriptions [19]. CHD is one of the most prevalent congenital malformations developing in the same period that nausea/vomiting presents. We used data from the HAVEN study and from a population selected from the Eurocat Northern Netherlands (NNL) registry to investigate the relationship between maternal use of antihistamines, prescribed for both pregnancy-related nausea/vomiting and other indications, and the risk of specific CHDs.
Methods Study populations HAVEN study The HAVEN study is a population-based, case–control family study to investigate lifestyles, environmental and genetic determinants in the pathogenesis and prevention of CHD [20]. HAVEN is a Dutch acronym for Heart Defects, Vascular status, Genetic factors and Nutrients. 361 eligible cases were identified at the age of around 16 months from the registries of four tertiary referral hospitals in Amsterdam, Leiden and Rotterdam, in the western part of the Netherlands. Two pediatric cardiologists trained at the same hospital diagnosed all the CHD phenotypes after birth by echocardiography and/or cardiac catheterisation and/or surgery. We assume that the different CHD phenotypes may be due to comparable exposures, but that they develop differently depending on genetic background and timeframe of the exposure. The phenotypes we included were: Tetralogy of Fallot (n = 44), complete atrioventricular
123
septal defects (AVSD) (n = 37), perimembranous ventricular septal defect (pVSD) (n = 98), aortic valve stenosis (n = 8), pulmonary valve stenosis (n = 60), coarctation of the aorta (n = 32), transposition of the great vessels (n = 50), hypoplastic left heart syndrome (n = 17), and miscellaneous types (n = 15) consisting of univentricular heart, aorta interruption, mitral valve atresia, tricuspid valve atresia, and corrected transposition of the great arteries. A genetic disorder or syndrome was present in 58/361 of the cases (16.0 %). 410 controls were randomly selected from the registries of the public child healthcare centres covering the source population of the cases. The Dutch healthcare system includes a regular check up of all newborns for health, growth and development by pediatric physicians in such a centre. The controls did not have any major malformations or chromosomal defects according to the medical records obtained at 16 months of age. They were invited to participate at the same age as the cases, at around 16 months. None of the children in the study population had been adopted. Participating families were not related to each other and were able to speak, read and write the Dutch language. We had 74.5 % response from the case families and 61.4 % from the control families. The main reasons not to participate were giving no permission to take a blood sample from their child for research purposes and the expected time effort. The study design has been described in more detail previously [20]. The Central Committee of Research in Humans and the Medical Ethical Committees of the participating hospitals approved the protocol and written consent from the parents was obtained prior to their participation. For the analyses, we included 361 cases with CHD and 410 controls between October 2003 and February 2007. Mothers with diabetes (n = 6), hyperhomocysteinaemia (n = 1), epilepsy (n = 6), hypertension (n = 6) and rheumatoid arthritis (n = 2) were excluded as these conditions could confound the associations we were investigating. We chose to study the mother–child pairs at 16 months after birth; this time of investigation was standardised and minimises recall bias and differential effects between the recall by case and control mothers regarding their periconceptional medication use, nausea/vomiting, and lifestyle behaviours. In addition, the study moment at more than 1 year after birth was chosen because most CHD are detected within the first year of life, thereby minimizing misclassification of control children. The mothers completed a questionnaire on demographics, illnesses, lifestyle behaviors and medication use at the study moment and during their periconceptional period, defined as 3 months prior to conception and up to 10 weeks after conception. The questionnaires were filled in at home and checked for
Results of two case–control studies
completeness and consistency by the researcher during the participant’s visit to our clinic. The history of prescribed and over the counter antihistamine use during the first 10 weeks of pregnancy was obtained from the questionnaire and coded according to the internationally accepted Anatomical Therapeutic Chemical (ATC) classification (controlled by the World Health Organisation Collaborating Centre for Drug Statistics Methodology) [21]. Medicines were coded based on their pharmacological and chemical properties. Anti-emetics were meclozine, meclozine/pyridoxine combination and cyclizine (piperazines), other antihistamines were cetirizine and levoceterizine (piperazines), amino-alkyl ethers and phenothiazines. We defined a child as having been exposed to the medicine during the first 10 weeks of pregnancy if the mother had taken the medicine during this 10-week period. In addition, the history of pregnancy-related nausea/vomiting was evaluated. Information on nausea/vomiting in the first 10 weeks of pregnancy was categorised as no nausea or vomiting, mild nausea with or without incidental vomiting, and severe nausea/vomiting. The last category was characterised by daily nausea with vomiting or with a serious influence on dietary intake. Mothers who reported any use of alcohol or cigarettes were considered as alcohol users and smokers. Periconceptional vitamin use was defined as daily intake during the whole periconceptional period, defined as starting 3 months prior to conception and up to 10 weeks after conception. Educational level was classified according to the definitions of Statistics Netherlands into low (primary/lower, vocational/intermediate, secondary), intermediate (intermediate, vocational/higher, secondary) and high education (higher vocational/university) [22]. Ethnicity was categorised as Western (both parents were born in the Netherlands, or in a European country or non-Western (one of the parents was of non-European origin) [22]. Standardised maternal measurements of weight (weighing scale, SECA, Germany, accurate to 0.5 kg) and height (anthropometric rod, SECA, Germany, accurate to 0.1 cm) were also performed during the clinic visit at 16 months after delivery. Body mass index (BMI) after birth was defined as weight in kilograms divided by the height in meters squared. Eurocat Northern Nederlands (NNL) Eurocat NNL is a population-based birth defect registry in the northern part of the Netherlands (the three Northern provinces Friesland, Groningen and Drenthe); it covers approximately 10 % of all Dutch births. A child can be registered with a defect up to the age of 16, there is no lower age limit. All types of births are included in the registry: live births, stillbirths, spontaneous abortions and terminations.
Parental informed consent is required to register a child and parents are asked to fill in a questionnaire on sociodemographic characteristics, prenatal screening methods and diagnostic tests, and prenatal exposure to possible risk factors (chemicals, drugs, etc.). With consent of the mother information on prescribed medications from the period 3 months before conception up to delivery are obtained from her pharmacy. The actual use of the prescribed medication and use of over the counter medication is later verified in a telephone interview. Birth defects are coded according to the ICD coding system. For births up to 2001, ICD-9 is used, while for births from 2002 onwards, ICD-10 is used [21]. Medication taken by the mother is coded according to the ATC classification system. From the Eurocat NNL database, 445 cases and 530 controls were selected. All the children were born in a 12-year period between 1st January 1997 and 31st December 2008. To comply with the HAVEN study, only live births were included. Of the 445 cases, 58 were diagnosed with Tetralogy of Fallot, 58 with complete AVSD, 121 pVSD, 14 with aortic valve stenosis, 39 with pulmonary valve stenosis, 51 with coarctation of the aorta, 66 with transposition of the great vessels and 38 with hypoplastic left heart syndrome. In total, 74/445 of the Eurocat cases (16.6 %) also had a genetic disorder. Since Eurocat NNL only includes children and fetuses with a birth defect, children with isolated hip dysplasia and/or luxation were selected as controls. The use of antihistamines has not been associated with these disorders. We excluded mothers with diabetes, hyperhomocysteinaemia, epilepsy, hypertension and rheumatoid arthritis because these conditions could confound the associations we were investigating. Anti-emetics were meclozine, meclozine/pyridoxine combination and cyclizine (piperazines), other antihistamines were cetirizine and levoceterizine (piperazines), amino-alkyl ethers and phenothiazines. Exposure to an antihistamine was defined as any antihistamine use during the first 14 weeks of pregnancy, calculated from the first day of the last menstrual period. No information was available on nausea/vomiting. Mothers who reported any periconceptional use of alcohol and/or cigarettes were considered as alcohol users and/or smokers. Periconceptional use of folic acid or multivitamins containing folic acid was defined as any use of these during the 3 months before conception up to the end of the first trimester. To record educational level, we used the definitions set by Statistics Netherlands. Ethnicity was categorised as ‘Western’ if the mother was born in Europe, North America, Australia, New Zealand or Indonesia, and ‘non-Western’ if the mother was born in any another country. Maternal weight and height
123
H. P. M. Smedts et al.
were based on the situation before pregnancy and were self-reported. Statistical analyses HAVEN study In the HAVEN analyses, normality of continuous variables was tested by the one-sample Kolmogorov–Smirnov Test. BMI showed a positively skewed distribution even after logarithmic transformation. Continuous variables of the maternal characteristics were therefore presented as medians with interquartile ranges, and differences between cases and controls were tested by the Mann–Whitney U test. Categorical variables were tested between cases and controls by the Chi squared test. We used a forward, stepwise, multivariable regression model to study associations between use of antihistamines during early pregnancy, stratified for anti-emetics and other antihistamines, and the risk of both overall CHD and the separate CHD phenotypes. The risk estimates were adjusted for the variables that were either significantly different between cases and controls or that had a P \ 0.1 in the logistic model (maternal age, educational level, ethnicity, parity, sex of the child and any medication except antihistamines). Crude and adjusted odds ratios (ORs) and 95 % confidence intervals (CIs) were calculated for use of antihistamines during early pregnancy if at least two cases had been exposed. In the HAVEN data, we then performed a combined analysis of anti-emetic use in early pregnancy and nausea/vomiting, with no antihistamine use and no nausea/ no vomiting as a reference, and separately presented the results for complete AVSD and pVSD defects, as these were found to be associated with antihistamine use. Trends across severity of nausea/vomiting towards CHD risk were evaluated by the linear-by-linear association test. Statistical analyses were performed with SPSS for Windows software (version 20.0; SPSS Inc, Chicago, IL, USA). Eurocat NNL The normality of continuous variables, maternal age and BMI was tested by the one-sample Kolmogorov–Smirnov Test. Maternal age was normally distributed for the Eurocat population. These characteristics were therefore presented as mean and standard deviation, and the differences between cases and controls were tested by the independent t test. Since BMI was not normally distributed, these characteristics were presented as medians with interquartile ranges and the differences between cases and controls were tested by the Mann–Whitney U test. Categorical variables were tested between cases and controls by the Chi squared
123
test. We used a logistic regression model to study associations between early pregnancy use of antihistamine, stratified for anti-emetics and other antihistamines, and the risk of both overall CHD and the separate CHD phenotypes. The risk estimates were adjusted for the same potential confounders or effect modifiers as in the HAVEN study. Crude and adjusted ORs and 95 % CIs were calculated if at least two cases had been exposed.
Results Socio-demographic and lifestyle characteristics of case mothers of a child with CHD and controls are presented in Table 1. HAVEN study case mothers showed a slightly higher age than controls, but ethnicity, educational level, BMI, parity, child gender and periconceptional exposures were not different except for a higher medication use in case mothers (27 %) than in healthy controls (19 %). For the Eurocat population case mothers were also slightly older than control mothers. The large majority of cases and controls were of Western origin, however the cases were slightly more often of non-Western origin (3 % [15/445] vs. 1 % [5/530]). Furthermore, we found a statistically significant difference in educational level and parity between cases and controls. There were more girls among the controls, since hip dysplasia is more common among girls. In the HAVEN study, antihistamine use was reported in 3.2 % of case and control mothers, of which 5.0 % in case mothers (18 of 361) and 1.7 % in control mothers (7 of 410) (Table 2). The use of antihistamines was associated with a threefold increased risk of CHD (crude OR 3.0, 95 % CI 1.2–7.3). Because of the small number of exposed cases, adjustment for potential confounders might not be useful. However, for completeness we added the adjusted ORs to Table 2. After stratification for anti-emetics, i.e., meclozine and the meclozine/pyridoxine combination as well as other antihistamines, the risk estimates were comparable although significance was lost due to small numbers. Stratification per CHD phenotype in the HAVEN study revealed that antihistamine use, particularly antiemetic medication, was associated with the occurrence of complete AVSD (crude OR 5.8, 95 % CI 1.03–32.8) and pVSD (crude OR 5.5, 95 % CI 1.4–20.7). Furthermore, other antihistamines were associated with an increased risk of Tetralogy of Fallot (crude OR 6.5, 95 % CI 1.05–39.8). Risk estimates for coarctation of the aorta, aortic valve stenosis, pulmonary valve stenosis, hypoplastic left heart syndrome, and the miscellaneous subgroup were not calculated as fewer than two cases in each category were exposed.
Results of two case–control studies Table 1 Characteristics of mothers of a child with congenital heart defects and of controls
HAVEN
Eurocat
CHD (n = 361)
Controls (n = 410)
CHD (n = 445)
Controls (n = 530)
32.9 (29.8–36.2)*
32.7 (28.8–35.1)
31.4 (4,543)^
30.0 (4,032)
24.4 (22.0–28.0)
24.4 (22.1–28.0)
23.5 (21.4–26.6)
23.5 (21.4–25.9)
300 (83)
318 (78)
422 (97)#
518 (99)
61 (17)
92 (22)
15 (3)
5 (1)
Low Intermediate
90 (25) 161 (45)
96 (24) 199 (48)
91 (21)* 211 (49)
75 (16) 280 (59)
High
Maternal age (years) 2
BMI (kg/m ) Ethnicity Western Non-Western Educational level HAVEN controls are healthy children without congenital malformation. Eurocat controls are children with an isolated hip disorder (dysplasia or dislocation). Data are median (interquartile range), mean (standard deviation) or n (%), compared by Mann–Whitney U Test or Chi squared test
110 (30)
115 (28)
126 (29)
124 (26)
Parity (first child)
152 (42)
200 (49)
147 (34)^
249 (48)
Sex of the child (male)
205 (57)
229 (56)
256 (58)^
78 (15)
Periconceptional use of 138 (38)
132 (32)
88 (21)
86 (17)
Cigarettes, yes
63 (18)
89 (22)
114 (26)
108 (21)
CHD congenital heart defect
B-vitamins, yes
182 (50)
206 (50)
300 (67)
367 (69)
*
Any medication, yes
98 (27)#
78 (19)
191 (43)
203 (38)
Antihistamines, yes
18 (5)#
7 (2)
22 (5)
20 (4)
P value B 0.05, # P value B 0.01, ^ P value B 0.001
Alcohol yes
In the Eurocat population, antihistamine use was reported in 4.3 % of case and control mothers, of which 4.9 % in case mothers (22/445) and 3.8 % in control mothers (20/530). The positive association between periconception use of antihistamine medication and complete AVSD was confirmed (OR 3.5, 95 % CI 1.4–8.7) in the Eurocat population. However, the association between periconception use of antihistamine medication and the total group of CHDs, pVSD and Tetralogy of Fallot was not confirmed: respectively we found OR 1.3, 95 % CI 0.7–2.5; OR 0.9, 95 % CI 0.3–2.6 and OR 0.9, 95 % CI 0.2–4.1. In the HAVEN study, in a subgroup analysis after excluding cases with a known genetic factor, the observed associations remained between antihistamine use and overall CHD risk (OR 3.2, 95 % CI 1.3–8.0), AVSD (OR 11.5, 95 % CI 1.2–111) and pVSD (OR 6.0, 95 % CI 2.0–17.7). In the Eurocat population, the association between antihistamines other than anti-emetics and AVSD also remained after excluding cases with a known genetic factor (OR 15.9, 95 % CI 1.4–184), but the association between overall antihistamine use and AVSD (OR 3.0, 95 % CI 0.7–14.0) attenuated to non-significant. In the HAVEN study, 52 % of case mothers and 59 % of control mothers reported nausea/vomiting. We demonstrated a significant trend towards a reduced overall CHD risk by increasing severity of nausea/vomiting, P = 0.020. Severe nausea and vomiting without the use of anti-emetic medication seemed to be associated with a reduced risk of CHD (OR 0.7, 95 % CI 0.5–0.98) (Table 3). However,
mothers with severe nausea/vomiting who used anti-emetic medication tended to have an increased risk of overall CHD offspring, albeit non-significant. Remarkably, the reduced risk in the group of mothers with severe nausea/ vomiting changed into an almost fivefold increased risk for particularly pVSD (OR 4.8, 95 % CI 1.1–21.8). These analyses could not be replicated as this information was not available for the Eurocat population.
Discussion Main findings The HAVEN study showed that antihistamine use was associated with a threefold increased overall CHD risk and a fivefold increased risk of pVSD and AVSD. In addition, a positive association was found with Tetralogy of Fallot for antihistamines other than anti-emetics. These associations were independent of the mother’s nausea/vomiting. In the Eurocat population we confirmed the association between the mother’s antihistamine use and the increased risk of having a child with AVSD, but we could not confirm the other associations. Strengths and limitations One of the strong points of our study is that we replicated the association with AVSD in two large, independent
123
123 3/7 8/7
Tetralogy of Fallot, n = 44 Transposition of the great vessels, n = 50
Complete AVSD, n = 37
Perimembranous VSD, n = 98
3/20 7/20 4/20
Tetralogy of Fallot, n = 58
Transposition of the great vessels, n = 66
Complete AVSD, n = 58
Perimembranous VSD, n = 121
3.0 (1.2–7.3)
0.9 (0.3–2.6)
3.5 (1.4–8.7)
1.2 (0.4–4.2)
0.9 (0.2–4.1)
1.3 (0.7–2.5)
5.1 (1.8–14.4)
5.1 (1.3–20.5)
2.7 (0.6–13.6) 2.4 (0.5–11.9)
2.6 (1.02–6.7)
0.7 (0.2–2.3)
3.6 (1.3–9.9)
0.9 (0.2–3.9)
0.4 (0.1–2.8)
1.2 (0.6–2.5)
5.2 (1.6 –16.7)
9.0 (1.6–50.8)
3.4 (0.6–19.9) 2.4 (0.4–14.2)
10/4
4/18
5/18
3/18
2/18
20/18
5/4
2/4
0/4 1/4
2.9 (0.9–9.3)
1.0 (0.3–2.9)
2.8 (1.0–7.8)
1.3 (0.4–4.7)
1.0 (0.2–4.5)
1.3 (0.7–2.6)
5.5 (1.4–20.7)
5.8 (1.03–32.8)
NE NE
0.8 (0.2–2.8)
2.8 (0.9–9.0)
1.3 (0.3–5.4)
0.5 (0.1–3.8)
1.2 (0.5–2.6)
4.5 (1.1–18.8)
6.8 (0.9–50.7)
NE NE
2.2 (0.6–7.2)
OR (95 % CI)*
0/2
2/2
0/2
0/2
2/2
3/3
1/3
2/3 1/3
8/3
Cases/ controls
NE
10,0 (1.4–72.4)
NE
NE
1.2 (0.2–8.6)
4.3 (0.9–21.6)
NE
6.5 (1.05–39.8) NE
3.1 (0.8–11.7)
OR (95 % CI)
Other antihistamines
NE
9.4 (1.0–85.2)
NE
NE
1.2 (0.1–10.1)
4.3 (0.8–23.7)
NE
8.5 (1.2–63.2) NE
2.8 (0.7–11.2)
OR (95 % CI)*
* Odds ratios adjusted to maternal age, and educational level, ethnicity, parity, sex of the child and periconceptional use of any medication except antihistamines. Anti-emetics are meclozine, meclozine/pyridoxine combination and cyclizine (piperazines). Other antihistamines are cetirizine and levoceterizine (piperazines), amino-alkyl ethers and phenothiazine
AVSD atrioventricular septal defect, VSD ventricular septal defect, CHD congenital heart defect, OR odds ratio, CI confidence interval, NE not estimated from logistic regression model
22/20 2/20
Total CHD, n = 445
Eurocat database, n = 445/530
18/7 2/7 2/7
Total CHD, n = 361
HAVEN database, n = 361/410
OR (95 % CI)
Cases/ controls
OR (95 % CI)*
Cases/ controls
OR (95 % CI)
Anti-emetics
Total antihistamines
Table 2 Early pregnancy exposure to antihistamines, stratified for anti-emetics and other antihistamines, and the risk of congenital heart defects
H. P. M. Smedts et al.
AVSD Atrioventricular septal defect, VSD ventricular septal defect, CHD congenital heart defect, OR odds ratio, CI confidence interval, NE not estimated from logistic regression model
* Odds ratios adjusted to maternal age, and educational level, ethnicity, parity, sex of the child, periconceptional use of any medication except antihistamines. The reference group comprised all mothers not exposed to antihistamine medication and who had no complaints of nausea and vomiting. We excluded 8 cases and 3 controls using antihistamines other than anti-emetics for this analysis
NE
4.8 (1.1–21.6) 5.4 (1.4–21.0) 5/4 6.8 (0.9–53.2) 5.2 (0.9–30.1) 2/4 1.8 (0.5–6.1) 10/4 ?
Severe
2.5 (0.8–8.2)
NE NE
NE 0/0
0/0 NE
NE NE
NE 0/0
0/0 NE
NE
0/0
NE
0/0
Mild ?
?
No
NE
1.0 (0.5–1.7)
1.0 (0.6–1.8) 1.0 (0.6–1.8)
0.8 (0.5–1.5) 23/117
28/118 1.0 (0.4–2.3)
0.9 (0.3–2.2) 0.7 (0.3–1.7)
0.9 (0.4–2.0) 10/118
8/117 0.7(0.5–1.01)
0.8(0.6–1.1) 0.8 (0.6–1.2)
Severe
0.7 (0.5–0.98)
96/118
80/117
Mild
-
Reference 39/168 Reference Reference Reference 167/168
(n = 36/407)
OR (95 % CI)* OR (95 % CI)
16/168
Reference
OR (95 % CI) Cases/controls
Cases/ controls (n = 353/ 407)
No -
-
Reference
OR (95 % CI)* OR (95 % CI) Cases/ controls (n = 95/407) OR (95 % CI)*
Perimembranous VSD Complete AVSD Total CHD Nausea/vomiting Anti-emetic use
Table 3 Maternal use of anti-emetics in combination with nausea and vomiting during early pregnancy in association with risk of congenital heart defect, HAVEN study results
Results of two case–control studies
populations with good information about medication use during pregnancy. In the HAVEN study, results from a standardised questionnaire were verified in a personal interview, while in the Eurocat registry, information on medication use is based on pharmacy records and verified in telephone interviews. Another strength of our study is the accuracy of the diagnoses of the specific CHD phenotypes: for the HAVEN study, two pediatric cardiologists diagnosed all the CHD cases, while for Eurocat, a medical doctor and clinical geneticist specializing in heart anomalies classified the cases and controls. One weakness of our study is that there could have been selection bias since only live births were included. CHD can be part of a chromosomal or genetic disorder and such pregnancies are terminated relatively more often or result in early fetal loss. Another type of bias that always has to be considered in case–control studies is recall bias [23]. In the HAVEN study we therefore standardised the data collection shortly after pregnancy. In Eurocat, the pharmacy data was for prescriptions dispensed shortly before and during pregnancy, and their use was verified with the mother. Use of anti-emetics was 5.0 and 5.4 % in the HAVEN and Eurocat populations, respectively, which agrees with the 5.8 % reported in a study on drug prescription patterns in the Netherlands [12]. Frequencies of use by controls were, however, lower at 1.7 and 3.9 %. No other medicines or maternal illnesses could explain the difference in overall medication use between cases and controls. If, in the control group, under-reporting is an issue, differential recall bias cannot be excluded in the HAVEN study. However, if there was selective recall bias, we would have expected mothers of a CHD child to recall more nausea/vomiting for which they used medication than the controls. The mothers were not aware of our specific questions on associations between medication use and CHD. In the HAVEN study, a fixed 2 year time interval was chosen between first trimester of pregnancy and the interview, which was similar between cases and controls. The distance between the region of Eurocat NNL and the Western part of the Netherlands where the HAVEN study cases were recruited is significant. Therefore, there is only a very small risk of overlap in patients between the two databases. Finally, we are aware that our observed associations are based on a small number of antihistamine exposures, leading to imprecise risk estimates with large confidence intervals. Interpretation Our findings are in line with the results reported by Queißer-Luft et al. [15] who showed that early pregnancy use of anti-allergics, mainly antihistamines, was associated with a seven- to nine-fold increased risk of CHD and
123
H. P. M. Smedts et al.
musculoskeletal anomalies. Our results might underestimate the true risk if antihistamines are also associated with cases of isolated hip dysplasia, which were used as controls in the Eurocat analyses. Recent data from the National Birth Defect Prevention Study also revealed positive associations between doxylamine, which is the major compound of BendectinÒ, and spina bifida, cleft lip, left ventricular outflow tract obstruction defects, and hypoplastic left heart syndrome [24]. Associations were also demonstrated between meclozine use and orofacial clefts, and between antihistamine use and pVSDs. In a prospective cohort study among 16,536 women exposed to meclozine and 540,660 unexposed women, there was no difference in the occurrence of congenital malformations. Although this was a large cohort, the data may be distorted by confounding by indication, since their results were not adjusted for nausea/vomiting [16]. In a second prospective cohort study, investigating frequencies of congenital malformations in 196 women with first trimester exposure to cetirizine and 1,686 controls, no increased risk was found for any congenital malformation, but the study might have been underpowered and medication among control mothers was not specified [17]. Furthermore, the number of exposed cases is too limited to detect only a modest teratogenic effect of antihistamine use. A meta-analysis revealed that antihistamine use in early pregnancy protected against all types of major congenital malformations [18]. However, by pooling all types of malformations, a differential effect for a specific birth defect would be considerably diluted. Moreover, they did not investigate confounding by indication. A meta-analysis of 16 cohort and 11 case–control studies on the association with BendectinÒ, which is pharmacologically closely related to meclozine, showed no difference in risk [25]. It is conceivable that the presence of nausea/vomiting indicates the adaptation of the maternal endocrine and metabolic system to pregnancy, and as such may protect the embryo from harmful or teratogenic exposures [26]. However, severe nausea/vomiting has been associated with orofacial clefting, renal dysgenesia and urinary tract defects [27].The absence of nausea/vomiting in pregnant women may also reflect a reduced production of placental human chorionic gonadotrophin and thyroxin, resulting in impaired placental growth and subsequently slower fetal growth and development [28]. Our results suggest that antihistamine medication might be a cardiac teratogen. An alternative hypothesis is that it is not the medication, but rather the resulting reduction in nausea/vomiting that increases CHD risk. A reduction in nausea/vomiting during early pregnancy might result in increased fetal exposure to harmful agents or unknown teratogens that stimulate the mother’s vomiting.
123
The underlying mechanisms for a potential teratogenic effect from antihistamines are unknown and we can therefore only speculate. In adults, blockage of H1-receptors, expressed in several tissues such as heart, placenta, and endothelium, can lead to reflex tachycardia, ventricular arrhythmias and hemodynamic changes in the heart [29]. H1-receptors play an important role in embryonic development [30]. If we hypothesise that the H1-receptors are also expressed in the fetal heart, its inhibition might lead to hemodynamic changes and subsequent CHD. AVSD is strongly correlated with Down syndrome [31]. In the HAVEN study, in total 21 of 33 cases with AVSD, and 2 of 3 exposed AVSD cases, also had Down syndrome. Down syndrome was not correlated with antihistamine use nor with complaints of nausea in the first trimester. In the Eurocat population, 35 of 58 cases with AVSD, and 4 of 7 exposed AVSD cases, had Down syndrome. We did not exclude cases with a genetic abnormality from our analyses, as this is clearly not the only factor determining risk for CHD. In this group, environmental exposures are also important modifiers. Furthermore, in our study population, there may be more cases with an underlying genetic factor, but they cannot be excluded because most of the genetic causes of CHD are not yet known. However, excluding cases with a known genetic factor did not substantially change the results.
Conclusion We found a positive association between maternal antihistamine use and risk of CHD, in particular pVSD and AVSD, in the HAVEN study. It appears that it is not the pregnancy-related nausea/vomiting, but rather the exposure to antihistamines that seems to increase the risk of CHD. The Eurocat data confirmed the association between antihistamine use and AVSD, but not with overall CHD risk; nor could we investigate the association with nausea/ vomiting. Our findings therefore warrant further investigation, which should take into account the need for a high number of exposed cases with a specific CHD phenotype. Acknowledgments We thank Professor W.A. Helbing, MD, PhD, Professor J. Ottenkamp, MD, PhD, and F.M.H. Siebel, MD, for their assistance in recruiting the case and control families. We are also grateful to Dr. M. Wildhagen, for data management and to Dr. A.C. Verkleij-Hagoort, Dr. L.M.J.W. van Driel and S. Borst for help in the data collection. We thank JL Senior for editing the final manuscript. The HAVEN study was financially supported by the Bo Hjelt Foundation (Grant 2005) and the Netherlands Heart Foundation (Grant 2002.B027). Eurocat NNL is funded by the Dutch Ministry of Health, Welfare and Sports. Conflict of interest
There are no conflicts of interest to declare.
Results of two case–control studies Ethical Standard The study protocol was approved by the Central Committee of Research in Humans in The Hague, the Netherlands, and by the local Medical Ethics Committees of the four participating hospitals (CCMO07.1052/MA/P03.0200; MEC212.508/2002/91).
References 1. Hoffman JI, Kaplan S. The incidence of congenital heart disease. J Am Coll Cardiol. 2002;39:1890–900. 2. Dolk H, Loane M, Garne E. Congenital heart defects in Europe: prevalence and perinatal mortality, 2000–2005. Circulation. 2011;123:841–9. 3. Yoon PW, Olney RS, Khoury MJ, Sappenfield WM, Chavez GF, Taylor D. Contribution of birth defects and genetic diseases to pediatric hospitalizations. A population-based study. Arch Pediatr Adolesc Med. 1997;151:1096–103. 4. Pierpont ME, Basson CT, Benson DW Jr, Gelb BD, Giglia TM, Goldmuntz E, et al. Genetic basis for congenital heart defects: current knowledge: a scientific statement from the American Heart Association Congenital Cardiac Defects Committee, Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics. Circulation. 2007;115:3015– 38. 5. Jenkins KJ, Correa A, Feinstein JA, Botto L, Britt AE, Daniels SR, et al. Non-inherited risk factors and congenital cardiovascular defects: current knowledge: a scientific statement from the American Heart Association Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics. Circulation. 2007;115:2995–3014. 6. Rosenquist TH. Folate, homocysteine and the cardiac neural crest. Dev Dyn. 2013;242:201–18. 7. Botto LD, Correa A. Decreasing the burden of congenital heart anomalies: an epidemiologic evaluation of risk factors and survival. Progr Paediatr Cardiol. 2003;18:111–21. 8. Steegers-Theunissen RP, Steegers EA. Nutrient-gene interactions in early pregnancy: a vascular hypothesis. Eur J Obstet Gynecol Reprod Biol. 2003;106:115–7. 9. Dally A. Thalidomide: was the tragedy preventable? Lancet. 1998;351:1197–9. 10. Leeder JS, Spielberg SP, MacLeod SM. Bendectin: the wrong way to regulate drug availability. Can Med Assoc J. 1983;129: 1085–7. 11. Einarson TR, Piwko C, Koren G. Quantifying the global rates of nausea and vomiting of pregnancy: a meta analysis. J Popul Ther Clin Pharmacol. 2013;20:e171–83. 12. Bakker MK, Jentink J, Vroom F, Van Den Berg PB, De Walle HE, De Jong-Van Den Berg LT. Drug prescription patterns before, during and after pregnancy for chronic, occasional and pregnancy-related drugs in the Netherlands. BJOG. 2006;113: 559–68. 13. Narrod SA, Wilk AL, King CT. Metabolism of meclizine in the rat. J Pharmacol Exp Ther. 1965;147:380–4. 14. King CT, Weaver SA, Narrod SA. Antihistamines and teratogenicity in the rat. J Pharmacol Exp Ther. 1965;147:391–8.
15. Queißer-Luft A, Eggers I, Stolz G, Kieninger-Baum D, Schlaefer K. Serial examination of 20,248 newborn fetuses and infants: correlations between drug exposure and major malformations. Am J Med Genet. 1996;63:268–76. 16. Kallen B, Mottet I. Delivery outcome after the use of meclozine in early pregnancy. Eur J Epidemiol. 2003;18:665–9. 17. Weber-Schoendorfer C, Schaefer C. The safety of cetirizine during pregnancy. A prospective observational cohort study. Reprod Toxicol. 2008;26:19–23. 18. Seto A, Einarson T, Koren G. Pregnancy outcome following first trimester exposure to antihistamines: meta-analysis. Am J Perinatol. 1997;14:119–24. 19. Markl GE, Strunz-Lehner C, Egen-Lappe V, Hasford J. Prescribing patterns of anti-emetic drugs during.pregnancy in Germany. Arch Gynecol Obstet. 2007;275:461–7. 20. Verkleij-Hagoort AC, Verlinde M, Ursem NT, Lindemans J, Helbing WA, Ottenkamp J, et al. Maternal hyperhomocysteinaemia is a risk factor for congenital heart disease. BJOG. 2006; 113:1412–8. 21. WHO Collaborating Centre for Drug Statistics Methodology. http://www.whocc.no/atcddd/. Accessed Sept 24 2008. http:// www.icd9data.com/ and http://www.icd10data.com/ICD10CM/ Codes. 22. Statistics Netherlands. Classification of educational level. Voorburg/Heerlen, the Netherlands, 2006. http://www.cbs.nl/menu/ methoden/classificaties/overzicht/soi/default.htm. Accessed Nov 27, 2007. 23. Cogswell ME, Bitsko RH, Anderka M, Caton AR, Feldkamp ML, Hockett Sherlock SM, et al. Control selection and participation in an ongoing, population-based, case-control study of birth defects: the National Birth Defects Prevention Study. Am J Epidemiol. 2009;170:975–85. 24. Gilboa SM, Strickland MJ, Olshan AF, Werler MM, Correa A. The National Birth Defects Prevention Study. Use of antihistamine medications and isolated major malformations. Birth Defect Res Part A. 2009;85:137–50. 25. McKeigue PM, Lamm SH, Linn S, Kutcher JS. Bendectin and birth defects: I. A meta-analysis of the epidemiologic studies. Teratology. 1994;50:27–37. 26. Flaxman SM, Sherman PW. Morning sickness: a mechanism for protecting mother and embryo. Q Rev Biol. 2000;75:113–48. 27. Czeizel AE, Puho´ E, Acs N, Ba´nhidy F. Inverse association between severe nausea and vomiting in pregnancy and some congenital abnormalities. Am J Med Genet A. 2006;140:453–62. 28. Huxley RR. Nausea and vomiting in early pregnancy: its role in placental development. Obstet Gynecol. 2000;95:779–82. 29. Simons FE. Advances in H1-antihistamines. N Engl J Med. 2004;351:2203–17. 30. Matsuyama K, Ichikawa T, Nitta Y, Ikoma Y, Ishimura K, Horio S, et al. Localized expression of histamine H1 receptors in syncytiotrophoblast cells of human placenta. J Pharmacol Sci. 2006; 102:331–7. 31. Marino B, Vairo U, Corno A, Nava S, Guccione P, Calabro R, et al. Atrioventricular canal in Down syndrome. Prevalence of associated cardiac malformations compared with patients without Down syndrome. Am J Dis Child. 1990;144:1120–2.
123
